CONCLUSIONS AND RECOMMENDATIONS

ERDC TN-DOER-C23

September 2001

Brunson et al. (1998) conducted a correlative type study designed, in part, to validate the use of the

oligochaete worm Lumbriculus variegatus as a bioaccumulation test species for freshwater sedi-

ments. Sediment samples and resident oligochaetes were collected from 13 navigational pools in

the Upper Mississippi and the St. Croix Rivers. Tissue residues were measured in L. variegatus

after 28 days of exposure to field-collected sediment and compared with tissue residues in

field-collected animals. While results of tissue analysis indicated that both PAHs and PCBs were

consistently found at concentrations above detection limits, the authors focused on the PAH data in

their evaluation of the 28-day L. variegatus bioaccumulation test. Lipid-normalized tissue concen-

trations of PAHs showed a positive correlation between laboratory-exposed and field-collected

organisms. Generally there was a change in tissue residue concentrations for individual PAHs; field

tissue residues were greater than laboratory tissue residues for the lower molecular weight PAH

tissue residues, and laboratory tissue residues were greater than field tissue residues for the high

molecular weight PAHs. The authors speculated that the observed differences between laboratory

and field-exposed animals may have been a result of any one of a number of factors. For example,

it is possible that the more water soluble lower molecular weight PAHs were lost during sampling

of sediments for the laboratory exposures. Differences may also have arisen from the spatial

heterogeneity of contaminants in the field (e.g., grab samples may have included sediment from

depths to which field-collected organisms were not exposed). Another possibility may be related

to the fact that field organisms may be exposed through a number of routes (e.g., sediment, water,

and food) whereas the primary route of exposure in laboratory bioaccumulation tests is through

direct contact and ingestion of sediment. There may also have been genetic/phenotypic or possibly

species differences between the lab and field organisms examined that may have lead to differences

in uptake. However, despite these differences, the authors indicate that the ratio of tissue concen-

trations for specific PAHs in laboratory and field-exposed organisms were generally similar

with about 90 percent of the corresponding concentrations for laboratory versus field-collected

organisms being within a factor of 2-3. In addition biota to sediment accumulation factors (BSAFs)

derived for lab and field data showed good correlation with the field-collected organisms, having

BSAF values equal to or higher than the laboratory (R2 = 0.80, p = 0.003). The good correlation

indicates that laboratory bioaccumulation tests with the oligochaete L. variegatus were good

predictors of bioavailability (Figure 8). Based on these findings the authors conclude, "Laboratory

results could be extrapolated to the field with a reasonable degree of certainty."

CONCLUSIONS AND RECOMMENDATIONS: Each of the two basic approaches employed

to date to field-validate laboratory bioassays (i.e., experimental/manipulative versus correlative/

observational) has inherent advantages and disadvantages. In general, correlative approaches are

subject to greater variability in the field portion of the assessment. In correlative studies, the spatial

variability of contaminants and contaminant mixtures remains intact. As one moves along the

"contaminant gradient," the type and amount of co-occurring contaminants change as do other

physical/chemical features of the sediment matrix and the biological communities that respond to

these factors. The difficulty in trying to ascribe differences in benthic community structure to

contaminant levels is evidenced by the low correlations found by Swartz et al. (1994) (Figure 4)

and McGee and Fisher (1999) (Figure 7). While spatial variability is inherent in environmental

contamination, such variability can confound attempts to validate responses measured in laboratory

tests where such variability is attenuated through homogenization of the sample matrix. The

experimental/manipulative approach attempts to overcome this variability via sample compositing

and/or homogenization to remove/reduce spatial variation.